Method of displaying the state of an artificial neural network

• US 5,222,210 A
• Filed: 12/03/1990
• Issued: 06/22/1993
• Est. Priority Date: 12/03/1990
• Status: Expired due to Fees

• Current Assignee: Motorola, Inc. (Motorola Solutions, Inc.)
• Original Assignee: Motorola, Inc. (Motorola Solutions, Inc.)
• Inventors: Leivian, Robert H.
• Primary Examiner(s): Harkcom, Gary V.
• Assistant Examiner(s): BURRASTON, NEIL
• Application Number:
  US07/620,944
• Time in Patent Office: 932 Days
• Field of Search: 395/12, 395/21, 395/60, 395/76, 395/22, 395/161, 395/160, 395/920, 364/489
• US Class Current: 715/854
• CPC Class Codes: G06N 3/105 Shells for specifying net layout

Abstract

A computer simulator is provided for displaying the state of an artificial neural network in a simplified yet meaningful manner on a computer display terminal. The user may enter commands to select one or more areas of interest within the neural network for further information regarding its state of learning and operation. One display mode illustrates the output activity of each neuron as representatively sized and shaded boxes within the border of the neuron, while another display mode shows the connectivity as weighted synapses between a user-selected neuron and the remaining neurons of the network in a similar manner. A third display mode provides a tuning curve wherein the synapses associated with each of the neurons are represented within the borders of the same. Both grid block and line graph type characterization are supported. The methodology allows large neural networks on the order of thousands of neurons to be displayed in a meaningful manner.

Claims

1. A method of displaying a state of a neuron network on a graphic display terminal with a plurality of neuron blocks, comprising the steps of:

positioning a cursor on the graphic display terminal with a user interface device;

receiving a command from said user interface device for determining XY-coordinates of said cursor on the graphic display terminal where said XY-coordinates of said cursor identifies a first one of the plurality of neuron blocks;

processing through second ones of the plurality of neuron blocks to identify a synaptic weight connection between said first one of the plurality of neuron blocks and a second one of the plurality of neuron blocks; and

displaying a box within said second one of the plurality of neuron blocks upon finding said synaptic weight connection between said first one of the plurality of neuron blocks and said second one of the plurality of neuron blocks where said box has an area representative of said synaptic weight connection.

2. The method of claim 1 further comprising the steps of:

retrieving output activity of the plurality of neuron blocks; and

drawing a second box within each of the plurality of neuron blocks representative of said output activity.